{"id":"https://openalex.org/W7204073147","doi":"https://doi.org/10.48550/arxiv.2608.21134","title":"Llama-Mobile: Efficient 2.7-Bit Quantization of VLMs","display_name":"Llama-Mobile: Efficient 2.7-Bit Quantization of VLMs","publication_year":2026,"publication_date":"2026-08-21","ids":{"openalex":"https://openalex.org/W7204073147","doi":"https://doi.org/10.48550/arxiv.2608.21134"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.21134","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21134","pdf_url":"https://arxiv.org/pdf/2608.21134","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.21134","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5080417728","display_name":"Luka Ribar","orcid":"https://orcid.org/0000-0002-7050-6387","observed_orcids":["https://orcid.org/0000-0002-7050-6387"]},"institutions":[{"id":"https://openalex.org/I4210139321","display_name":"Graphcore (United Kingdom)","ror":"https://ror.org/03b3h3q73","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210139321"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Ribar, Luka","raw_affiliation_strings":["Graphcore Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graphcore Research","institution_ids":["https://openalex.org/I4210139321"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147982623","display_name":"Jeevan Bhoot","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bhoot, Jeevan","raw_affiliation_strings":["Arm"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Arm","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5126345015","display_name":"Douglas Orr","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210139321","display_name":"Graphcore (United Kingdom)","ror":"https://ror.org/03b3h3q73","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210139321"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Orr, Douglas","raw_affiliation_strings":["Graphcore Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Graphcore Research","institution_ids":["https://openalex.org/I4210139321"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9940999746322632,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9940999746322632,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.0012000000569969416,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.00039999998989515007,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/vision-language-models","display_name":"vision-language models","score":0.972000002861023},{"id":"https://openalex.org/keywords/model-compression","display_name":"model compression","score":0.5640000104904175},{"id":"https://openalex.org/keywords/mobile-deployment","display_name":"mobile deployment","score":0.6809999942779541},{"id":"https://openalex.org/keywords/visual-question-answering","display_name":"visual question answering","score":0.8149999976158142},{"id":"https://openalex.org/keywords/resource-constrained-devices","display_name":"resource-constrained devices","score":0.7020000219345093}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7159000039100647},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.6348000168800354},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4948999881744385},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.4659000039100647},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.44920000433921814},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41609999537467957},{"id":"https://openalex.org/C186967261","wikidata":"https://www.wikidata.org/wiki/Q5082128","display_name":"Mobile device","level":2,"score":0.4120999872684479},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3880999982357025},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3880999982357025},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.37220001220703125},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.32330000400543213},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3140999972820282},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.31290000677108765},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.29670000076293945},{"id":"https://openalex.org/C3018263672","wikidata":"https://www.wikidata.org/wiki/Q1296251","display_name":"Efficient algorithm","level":2,"score":0.28519999980926514},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.2800000011920929}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.21134","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21134","pdf_url":"https://arxiv.org/pdf/2608.21134","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.21134","doi":"https://doi.org/10.48550/arxiv.2608.21134","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.21134","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.21134","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.21134","pdf_url":"https://arxiv.org/pdf/2608.21134","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204073147.pdf","grobid_xml":"https://content.openalex.org/works/W7204073147.grobid-xml"},"referenced_works_count":22,"referenced_works":["https://openalex.org/W2108598243","https://openalex.org/W2242818861","https://openalex.org/W3120043490","https://openalex.org/W4213213306","https://openalex.org/W4307934016","https://openalex.org/W4387725429","https://openalex.org/W4401306886","https://openalex.org/W4402670692","https://openalex.org/W4405903187","https://openalex.org/W4406051710","https://openalex.org/W4410636396","https://openalex.org/W4413146209","https://openalex.org/W4414786525","https://openalex.org/W4414939355","https://openalex.org/W4416780398","https://openalex.org/W4417006854","https://openalex.org/W6947991961","https://openalex.org/W6948369276","https://openalex.org/W7077077575","https://openalex.org/W7126071242","https://openalex.org/W7166871762","https://openalex.org/W7196965917"],"related_works":["https://openalex.org/W4416235597","https://openalex.org/W4416258439","https://openalex.org/W4417099704","https://openalex.org/W7125376043","https://openalex.org/W7077077575","https://openalex.org/W4390437694","https://openalex.org/W7160255335","https://openalex.org/W4402701845","https://openalex.org/W7167884527","https://openalex.org/W4416157564"],"abstract_inverted_index":{"Deploying":[0],"vision-language":[1],"models":[2],"(VLMs)":[3],"on":[4,27,62,87],"mobile":[5],"devices":[6],"is":[7],"challenging":[8],"due":[9],"to":[10,41,50,78],"their":[11],"significant":[12],"memory":[13],"and":[14,45],"compute":[15],"requirements.":[16],"We":[17,65],"present":[18],"a":[19,33,55,88],"framework":[20],"for":[21,24],"quantizing":[22],"VLMs":[23],"efficient":[25,60],"inference":[26],"resource-constrained":[28],"hardware.":[29],"Our":[30],"approach":[31,68],"combines":[32],"quantization":[34],"pipeline":[35],"that":[36],"uses":[37],"the":[38,51,71],"model":[39,77],"itself":[40],"generate":[42],"training":[43,52],"data":[44],"does":[46],"not":[47],"require":[48],"access":[49],"setup,":[53],"with":[54,81],"novel":[56],"2.7-bit-per-parameter":[57],"format":[58],"supporting":[59],"execution":[61],"Arm":[63],"CPUs.":[64],"validate":[66],"our":[67],"by":[69],"compressing":[70],"Llama":[72],"3.2":[73],"11B":[74],"Vision":[75],"Instruct":[76],"3.7":[79],"GB":[80],"8-bit":[82],"activations,":[83],"preserving":[84],"strong":[85],"performance":[86],"set":[89],"of":[90],"standard":[91],"visual":[92],"question":[93],"answering":[94],"tasks.":[95]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-25T00:00:00"}